Subtitling is a specific field of translation which combines the art of translation with space and time constraints. Subtitle creators must be able to convey the main idea while giving viewers enough time to watch the visual content. They also have to follow some technical specifications.
As a standard, we deliver subtitles in an SRT file, which is a plain text file containing segmented translation and timing of the subtitles. The text is encoded in universal UTF-8 format that supports curly quotation marks and apostrophes, ellipses and other common punctuation. We follow the typography rules of a respective language. Also, we recommend that you always check in advance whether your media player supports this encoding format. If not, we can replace punctuation marks with displayable alternatives (three dots instead of an ellipsis, straight quotes instead of curly quotes, etc.) or select different encoding.
Constraints for the subtitles are both spatial and temporal. Our standard is to have one or two lines per subtitle with a maximum of 42 characters per line (including spaces) for 16:9 videos. We use 38 characters for older 4:3 video formats and 35 characters for 9:16 social media formats. These parameters ensure that the subtitles do not take up too much space while still being easily legible.
Temporal constraints have to do with the ability of viewers to read text. This ability is expressed in reading speed measured in characters per second (CPS). Our standard is a maximal CPS of 14–17 depending on the video type and its intended audience. We can go for a higher speed in non-narrative programmes, such as podcasts or interviews, that have limited or no action on screen. Display time of the subtitles is also important. The minimum duration is 1.5 seconds, the maximum one is 6 seconds. With longer display time, viewers tend to read the text twice.
The quality of the translation and the ability to convey the main idea is crucial when creating subtitles. Our standard is high quality translation guaranteed by our experienced and well-educated linguists. We also offer proofreading by native speakers.
In recent years, there has been an increased use of machine or AI-assisted translation in the translation industry. For the time being, our company has not been implementing these tools since the quality of their outcomes fails to be fully satisfying. Audiovisual translation is unique in many ways. Translators are supposed to reflect and convey context, both verbal and non-verbal expressions, style, local nuances, humour, etc.
If you want to create subtitles from your own machine translated text, we can do this with no interventions to the text, however, we are obliged to address their possible shortcomings. Please be aware that the process of the machine translation post-editing may take the same, or even greater, amount of time than translating the source text from scratch. This fact is reflected in our rates for such services. We do not reject machine translation completely, AI-supported processes definitely have some potential in the translation industry, but, as mentioned earlier, we are fully aware of the deficiencies in which they may result.
Another important thing is the visual aspects of subtitles. Our subtitles are large enough to read comfortably. We use easily legible fonts and take pride in our good typography. The default colour of the subtitles is white. However, we can always change the font and other parameters based on your preferences.
Subtitle creation is not only a business but also a hobby. We also run a non-profit website o.titulkovani.cz that covers the issue at a greater depth (in Czech).